HP StorageWorks 2/128 HP StorageWorks SAN Director 2/128 Fabric OS V4.2.x Rele - Page 15

Fabric OS: switch, Fabric OS: commands

Page 15 highlights

Important Notes Table 5: Fabric OS Area Information (Continued) Fabric OS Area Fabric Device Management Interface (FDMI) Fabric OS: CLI commands, failover, and port disable Fabric OS: commands Fabric OS: switch beaconing Fabric OS: switch reboot and blade repair Description An HBA will be allowed to register even though the originating port is not on the HBA's registered port list. This is intended behavior, included to test error cases. Changing port configurations during a failover might cause ports to be disabled. Reissue the command after the failover is complete to bring the ports online. Under the root account, issuing Fabric OS commands in parallel through scripts could cause the kernel task to consume excessive memory. Workaround: When using scripts to issue Fabric OS commands, wait for one command to finish before issuing another. Switch beaconing is not preserved across a failover. If you start beaconing, a failover causes all lights to stop flashing. Workaround: If this occurs, reissue the command to resume switch beaconing. Switch reboot will fail in the SAN Director 2/128 if there are faulty port blades. Note: Verify that all blades are in working order before performing a switch reboot. Switch reboot is meant to be issued after all repairs are complete. If you perform a switch reboot and find a faulty blade, remove the blade and reboot will continue. Fabric routing, Fabric Manager: domain overlap Fabric Watch: e-mail alert error message Firmware download Workaround: Identify and remove the faulty blade, using the slotshow command. Issuing a configdefault command followed by reboot or switch disable or enable can cause the fabric to segment due to possible domain overlap. Workaround: Before rebooting the fabric, ensure that all switches are properly configured to avoid domain overlap between the logical switches. If an event occurs while Fabric Watch e-mail alerts are being enabled, the message ErrLog: Error Level=3 [(null)] is captured to the system error log. This message is from SMTP and can be ignored. During a firmware download, rebooting or power cycling the CPs could corrupt the compact flash. CAUTION: Do not attempt to power off the CP board during firmware download, to avoid high risk of corrupting your flash. SAN Director 2/128 Fabric OS 4.2.x Release Notes 15

  • 1
  • 2
  • 3
  • 4
  • 5
  • 6
  • 7
  • 8
  • 9
  • 10
  • 11
  • 12
  • 13
  • 14
  • 15
  • 16
  • 17
  • 18
  • 19
  • 20
  • 21
  • 22
  • 23
  • 24

Important Notes
15
SAN Director 2/128 Fabric OS 4.2.x Release Notes
Fabric Device
Management Interface
(FDMI)
An HBA will be allowed to register even though the originating port is not
on the HBA's registered port list. This is intended behavior, included to test
error cases.
Fabric OS: CLI
commands, failover,
and port disable
Changing port configurations during a failover might cause ports to be
disabled. Reissue the command after the failover is complete to bring the
ports online.
Fabric OS: commands
Under the root account, issuing Fabric OS commands in parallel through
scripts could cause the kernel task to consume excessive memory.
Workaround:
When using scripts to issue Fabric OS commands, wait for
one command to finish before issuing another.
Fabric OS: switch
beaconing
Switch beaconing is not preserved across a failover. If you start beaconing,
a failover causes all lights to stop flashing.
Workaround:
If this occurs, reissue the command to resume switch
beaconing.
Fabric OS: switch
reboot and blade
repair
Switch reboot will fail in the SAN Director 2/128 if there are faulty port
blades.
Note:
Verify that all blades are in working order before performing a
switch reboot. Switch reboot is meant to be issued after all repairs are
complete. If you perform a switch reboot and find a faulty blade, remove the
blade and reboot will continue.
Workaround:
Identify and remove the faulty blade, using the
slotshow
command.
Fabric routing, Fabric
Manager: domain
overlap
Issuing a
configdefault
command followed by reboot or switch disable
or enable can cause the fabric to segment due to possible domain overlap.
Workaround:
Before rebooting the fabric, ensure that all switches are
properly configured to avoid domain overlap between the logical switches.
Fabric Watch: e-mail
alert error message
If an event occurs while Fabric Watch e-mail alerts are being enabled, the
message
ErrLog: Error Level=3 [(null)]
is captured to the
system error log. This message is from SMTP and can be ignored.
Firmware download
During a firmware download, rebooting or power cycling the CPs could
corrupt the compact flash.
CAUTION: Do not attempt to power off the CP board during firmware
download, to avoid high risk of corrupting your flash.
Table 5:
Fabric OS Area Information (Continued)
Fabric OS Area
Description